Open swvanderlaan opened 6 years ago
An .sh master script with a parameter for file for the pathway for each cohort (so kind of like the .filest.list) and then a parameter for where the lambda gets calculated, a parameter for the N column.
I'll try this next.
Add in params-file generator (cohort-name, lambda [after QC], avg. sample size, beta-correction factor) 🚧 If you have many cohorts (10+) or many (sub-)meta-analyses creating the param-file by hand can be a nuisance. We should make a perl/python script (probably fastest) to print out these things in one go...